COUNTRY REPORT ON SUPPORT FOR STATISTICS

CRESS

CAMBODIAJUNE 2017

© Christopher Rose

Country Report on Support to Statistics:

Cambodia

June 2017

Acknowledgments

The first Country Report on Support to Statistics (CRESS) Cambodia was prepared by the National Institute of Statistics (NIS) in collaboration with the Cambodia Statistics Coordinating Committee (SCC) through the leadership and guidance of Mr. San Sy Than, Secretary of State for Statistics, Ministry of Planning, and Mrs. Hang Lina, Director General of NIS, and Chairperson of the SCC. The CRESS project was made possible through the assistance and support of the Partnership in Statistics for Development in the 21st Century (PARIS21).

The CRESS Cambodia Country Team of NIS headed by Mr. Lay Chhan, Deputy Director General, supported by Mr. Chan Thy Samrith and Mr. Som Bony, Bureau Chief and Deputy Director, respectively, facilitated the data gathering, stakeholder consultations and validation. PARIS21 CRESS Team with Mr. John Frederick de Guia, international consultant, Mr. So Sovannarith, national consultant, and Ms. Millicent Gay B. Tejada, Regional Programme Co-ordinator for Asia and the Pacific, provided the direction and technical guidance of the CRESS project.

Acknowledgments are due to the honorable ministers and the heads and staff of the statistical units of the various line ministries and selected agencies of the Royal Government of Cambodia as well as the focal persons of the responding international development organizations in Cambodia for their active cooperation and valuable contribution.

Special thanks to the following for their valuable contribution in the project: General Department of Identification, Ministry of Interior; Hak Sophal of the Statistics Department, Ministry of Economy and Finance; the Department of Agricultural Statistics, Ministry of Agriculture, Forestry and Fisheries; Ma Sovanna of the Planning and Statistic Office, Department of Planning and Public Relations, Ministry of Rural Development; Moeun Ratanak of the Department of Planning Statistics and Trade Information, General Department of Domestic Trade, Ministry of Commerce; Pong Pitin of the Education Management Information System, Ministry of Education, Youth and Sport; Toeng Dara and Yim KomSan of the Department of Planning, Statistics and Sum-up, Ministry of Social Affairs, Veteran and Youth Rehabilitation; Neang Nim of the Statistics Department, Ministry of Land Management, Urban Planning and Construction; Department of Planning, Ministry of Environment); Ork Tola and Huot Daro of the Department of Accounting Planning and Administration, Ministry of Information; Nen Ekneta of the Department of Planning and Statistic, Ministry of National Assembly-Senate Relations and Inspection; Dr. Chivon of the Department of Health Information System, Ministry of Health; Horn Rithy of the Department of Planning and Statistic, Ministry of Culture and Fine Arts; Kong Sopheareak of the Statistics and Tourism Information Department, Ministry of Tourism; Te Vouchlim, Tim Phiyary, and Khim Sovanny of the Planning and Statistics Department, Ministry of Women’s Affairs; Moeung Sathya and Song Makara of the Strategic Planning and Policy Department, State Secretariat of Civil Aviation; Pich Sorya and Sem Vannak of the Department of Financial planning and Supply, Ministry of Civil Service; the Statistics Department (Balance of Payment Statistics Division, Monetary and Financial Statistics Division, and External Sector Survey Division), National Bank of Cambodia; and Ing Sokun of the General Directorate of Planning and Sem Simanith of the Department of Community Development Plan, General Secretariat for Population and Development of the Ministry of Planning; and Piseth Meng of the Department of Foreign Affairs and Trade, Australia; Sarany Khoeng of the Embassy of Sweden; Chea Vibol of the United Nations Children's Fund; Sok Lang of the United Nations Development Programme; Soktha Yi of the United Nations Population Fund; Sok Narom of the World Bank; and Mo Mai of the World Health Organization.

i

Message from the Senior Minister, Minister of Planning

The preparation of the Country Report on Support to Statistics (CRESS) Cambodia is indeed timely and relevant as Cambodia finally begins the formulation of the National Strategy for the Development of Statistics to succeed the Statistics Master Plan of 2008-2015. The CRESS Cambodia will be a useful tool in understanding how the different line ministries and agencies finance their statistical operations, capacity development, and data quality improvement. The Cambodia national statistical system is currently facing a significant financing gap as several major censuses and surveys are expected to be conducted in the next three to five years, such as the 2019 Population Census, 2020 Agricultural Census, annual socio-economic surveys, among others. As financial support from both the government and international development partners continue to decline, I hope that the CRESS Cambodia can help raise awareness on the need to invest more in statistics to ensure sustained production of critical data in support of the National Strategic Development Plan and the Sustainable Development Goals of Cambodia. The adoption and regular dissemination of the CRESS Cambodia is very important for statistical advocacy towards increased investment in statistics which will ensure that the goals of the Cambodia national statistical system for more efficient and effective implementation of statistical policies, among others, are fully met. I urge the members of the Statistics Advisory Council, the line ministries and agencies, especially the designated statistical units, and the international development community to make use of the CRESS Cambodia as reference for improving expenditure planning towards more rational partnerships and investment in statistics. Finally, I thank PARIS21 for their unwavering support for the continued development of institutional and technical capacities in the Cambodia national statistical system, especially NIS.

H.E. Chhay Than Senior Minister, Minister of Planning

ii

Message from the Director General of the National Institute of Statistics

The CRESS Cambodia is critical to the preparation of two (2) key strategic policy documents in the country, the National Strategy for the Development of Statistics (NSDS) and the National Strategic Development Plan (NSDP). The CRESS Cambodia will help identify gaps in financing of statistics not only in terms of operations, technology, and human resource development but also in terms of the priority sectoral or subject matter areas in support of the monitoring of the NSDP and Sustainable Development Goals. While the initial CRESS Cambodia has provided many lessons for improvement, there remain several challenges to statistics such as the independence of statistical budgetary and financial management, need for more well-trained statistical staff, and data quality issues which are critical in ensuring the production of statistics for evidence-based decisionmaking and development monitoring. The lessons generated by the CRESS Cambodia initiative should provide critical inputs for more open and productive dialogue within the government ― between designated statistical units and high level officials, between NIS and the designated statistical units, and between the Ministry of Economy and Finance and line ministries and agencies, and between government and the international development partners, towards greater support and increased investment in statistics to support the NSDP and SDGs. I encourage the members of the Statistics Coordinating Committee to continue our cooperation and collaboration by making full use of CRESS Cambodia and by sustaining its preparation to help monitor progress in the financing of priority statistics in the country. With the guidance of the Statistics Advisory Council, I am confident that we will achieve our shared goal of making statistics count in our country’s development. On behalf of NIS, I would like to express my appreciation to PARIS21 for their continued advice and commitment to helping the Cambodia national statistical system, especially NIS, improve capacities in undertaking strategically important statistical initiatives such as the CRESS and the NSDS. I would also like to thank the CRESS Cambodia Country Team from NIS and PARIS21 for their valiant efforts in completing the CRESS Cambodia.

H.E. Mrs. Hang Lina Delegate of the Royal Government of Cambodia in charge of

Director General of National Institute of Statistics

iii

Table of Contents Messages

Message from the Senior Minister, Minister of Planning .................................................................................. i Director General of the National Institute of Statistics ..................................................................................... ii

Executive Summary .............................................................................................................................................. iii 1. Introduction ....................................................................................................................................... 1 The Country Report on Support to Statistics ....................................................................................... 2

About PARIS21 ......................................................................................................................................... 2 2. Why CRESS Cambodia .................................................................................................................................... 3 3. The CRESS Cambodia ..................................................................................................................................... 5 3.1 Methodology ..................................................................................................................................... 5 3.1.1 Organizing the CRESS Cambodia Country Team ............................................................................ 5

3.1.2 Launching forum ............................................................................................................................ 5 3.1.3 Scope of the data .......................................................................................................................... 6 3.1.4 Temporal coverage ....................................................................................................................... 7 3.1.5 Data sources ................................................................................................................................. 7 3.1.6 The instruments ............................................................................................................................ 8 3.1.7 Data collection .............................................................................................................................. 8 3.1.8 Data validation .............................................................................................................................. 9

4. The CRESS Cambodia results and findings ...................................................................................................... 10 About the data ....................................................................................................................................... 10

4.1 Analysis of support to statistics ........................................................................................................ 10 4.1.1 Support from domestic (government) sources, 2013-2016 ....................................................... 11

4.1.2 Support from external sources, 2013-2016 ................................................................................ 16 As reported by the government .......................................................................................... 16 As reported by development partners ................................................................................ 19

4.1.3 Financing prospects for 2017-2019 ............................................................................................ 22 4.2 Limitations ....................................................................................................................................... 24 4.3 Challenges ........................................................................................................................................ 25

5. Lessons ......................................................................................................................................................... 29 6. Recommendations ....................................................................................................................................... 30

Appendix

Appendix A – List of tables Appendix B – List of figures Appendix C – Results tables Appendix D - Respondents Information Appendix E – Acronyms Appendix F – List of designated official statistics Appendix G – References

Annex Annex A – CRESS Cambodia Concept Note Annex B – Launch programme, list of participants, and presentation materials Annex C – Modified UNECE Classification of Statistical Activities Annex D – CRESS Cambodia Questionnaires Annex E – CRESS Cambodia Database Annex F – The CRESS Cambodia Country Team

iv

Executive Summary The Country Report on Support to Statistics (CRESS) Cambodia aims to provide insights on the extent, trends, and dynamics of financing statistics in the country. It seeks to inform stakeholders on the sources of financing, the priority areas of investment, and expenditure plans for statistics to initiate dialogue on appropriate solutions to financing gaps. Ultimately, CRESS aims to help in advocacy for increased investment in statistics and more rational expenditure planning for the national strategies for the development of statistics (NSDS) in support of the sustainable development goals (SDGs) and other national development outcomes. The report also provides recommendations for improving the measurement and monitoring of support to statistics. The CRESS Cambodia is based on the basic methodology developed by PARIS21 and drew from the experience of the 2015 CRESS pilot study in the Philippines. The preparation of the CRESS came at an appropriate time as the government begins the process towards the formulation of the NSDS that will build upon the gains of the Statistics Master Plan 2008-2015 and address persistent and emerging statistical issues. The CRESS Philippines methodology and instruments were adapted to consider the situation in Cambodia. The questionnaires were modified for better clarity and expanded to collect additional details especially on external funding received by government. The questionnaires were translated into Khmer, the local language, to further facilitate response. The target respondents included the National Institute of Statistics (NIS), the Ministry of Planning, and the twenty-seven ministries and agencies that are mandated by the Statistics Law of 2015 to produce and report designated official statistics. Twenty of the target government respondents submitted filled out questionnaires. On the other hand, seven of fifteen international development partners invited to participate in the study responded and returned either filled out questionnaires or reference documents. Other sources of information were tapped to supplement or validate responses. The survey did not cover support from the private sector, non-government organizations, academic and research institutions, and local government units. The study learned that the total financing for statistics as reported by twenty ministries and agencies amounted to US$ 35.8 million for the four-year period 2013-2016. About seventy-three percent of funding (or US$ 26.3 million) was provided by the government while twenty-seven percent (US$ 9.5 million) was received from external sources. Financing for statistics, both from government and external sources, appears to have been on the decline during the period 2013-2016. Data shows that total funding for statistics dropped from US$ 12.9 million in 2013 to US$ 7.8 million in 2016. The general downtrend can be attributed to the decrease in funding for major national censuses and surveys which are mostly periodic or conducted in multiyear intervals. With the significant drop in external financing, the share of government funding increased from 56.6 percent in 2013 to about 86.5 percent in 2015 and 2016. Total regular budget allocation for statistics decreased continuously from US$ 5.7 million in 2013 to US$ 2.9 million in 2016. It accounted for, on average, 74 percent of total government allocation for statistics while budget for special and nonrecurring statistics projects, including government counterpart funding for activities with external support, comprised about 26 to 27 percent during the period.

v

Results show a steep decline in government allocation to NIS, the primary statistical agency in Cambodia, from 2013 to 2016 which is a marked contrast to the increasing trend of government budget for statistics from 2005 to 2013 based on data from a separate budget monitoring report compiled by NIS. Similarly, external assistance for NIS, which had been mainly allocated for the conduct of major national censuses and surveys, continued to decrease significantly during the same period. Meanwhile, responses from seven international development partners indicate a total of US$ 21.3 million worth of financial and technical assistance provided to specific government ministries and agencies for the implementation of various statistics projects during the period 2013-2016. Such amount however is US$ 11.8 million more than the level of external funding reported by the responding ministries and agencies. Six of the seven responding development partners have indicated continued support for various statistics projects of the government in 2017-2019 with a total cost of US$ 8 million. The study has several limitations. The limited coverage of institutions and relatively low response rate together with incomplete data may have in general resulted in the underestimation of financing of statistics, both from domestic and external sources. There were observed disparities between responses of the responding government and development partner institutions. As most of the participating ministries and agencies do not have a specific line item for statistics in their annual budget, there was difficulty in identifying and classifying statistical activities and resources and the corresponding budget or expenditure. The lack of details and the difficulty in the classification of statistical activities prevented critical analysis of financing of designated official statistics and indicators for monitoring the national strategic development plan and SDGs which could have made the results more meaningful for policy. If the CRESS is to become an effective tool for monitoring investment in statistics in Cambodia, there needs to be increased capacity, greater priority, and more concrete strategic directions in managing the process. The Statistics Coordinating Committee (SCC) with support of NIS needs to institute appropriate measures that promote better coordination, greater access to budget data and information, and increased technical capacity within the ministries and agencies. The CRESS will likewise benefit from strengthened coordination and information sharing between the national statistical system (NSS) and development partners. All these mechanisms will contribute to effectively addressing many of the identified financing data quality issues. The CRESS may be also considered as one of the mechanisms of a wider monitoring and evaluation (M&E) plan for statistics that should include the monitoring of the NSDS and the NSDP and SDG indicators. The Cambodia NSS may have found themselves at a crossroads where external financing is diminishing and more local resources are increasingly needed to sustain statistical outputs. NIS, as the coordinator of the NSS, will need to urgently step up efforts to institutionalize expenditure planning for statistics, perhaps through the new government policy on program budgeting, towards the creation or inclusion of a separate line item for statistics and increased allocation in the Budget Law. There is strong impetus for a more rational, well-coordinated, and results-based approach to financing statistics to ensure that the most relevant and priority indicators and statistics are produced in support of the NSDP and SDGs and their ultimate beneficiaries – Cambodia and her people.

1

Country Report on Support to Statistics Cambodia 1. Introduction Statistics plays a critical role in development, a view shared by many a planner, policy maker, researcher, data analyst, and statistician in national and international institutions as well as among academics. Collecting quality data and statistics is a hard task and expectedly more difficult in countries where resources are limited. While governments and international development agencies continually invest in improving capacities of statistical systems to produce better quality statistics, there have not been many studies of how much is essentially being invested and whether that investment is adequate to generate statistics that matters most to users. Recent decades saw an astonishing growth in information technology triggering a virtual explosion of data, information, and knowledge, good and bad, about anything and everything from everyone, everywhere. The world has become truly global, connected by complex networks of information expressways only the most visionary could ever imagine or predict would become reality. Information technology has permeated every single vein of society influencing many of the ways mankind interact and live. Social media may be the closest model of this intertwined virtual-real world. It seems anyone or anything is now a mere tweet or post, a like or hashtag but which brings about new data, information, and knowledge. Unfortunately, social media have also given rise to the unwanted lies, fake news, and the ingenuous ‘alternative facts’, as if real facts have alternatives. What does all this have to do with statistics? A lot, a huge lot. For one, there is enormous data out there and yet data remains wanting especially the kind that agents of change --- people, governments, civil society, and international development institutions, depend so much upon to make a difference. The concept of data revolution is indeed revolutionary. The data revolution may be seen in two perspectives. One, it signals a much needed and urgent transformation of conventional policies and business processes in statistics towards an environment that allows for the creation and generation of data by everyone -- people, institutions, and systems, for their own and/or others’ use in informed decision making. The other perspective is that it is a revolution where data and statistics are based on real facts, not faked, not alternatives, and are used to effect positive change for improved welfare and sustainable development. Developing data ecosystems that optimize many different data sources and resources to generate statistics that is useful, inclusive, and supportive of development work may just be the best way to arresting and narrowing data and development gaps. The adoption of the Sustainable Development Goals (SDGs) ushers in a wide range of both challenges and opportunities to make a better world. In many countries, however there are simply more challenges than opportunities. In these countries, including Cambodia, a balanced mix of strategic and practical solutions is essential in creating new and optimizing existing resources to address persistent societal problems and emerging development issues. Key to ensuring that development initiatives are on track towards the desired goals is effective results-focused monitoring and evaluation, one that is based on sound indicators and quality statistical information. Statistics thus remains a critical factor in development but now it has also become an important goal in development.

2

As the countries adopt the SDGs, the need to strengthen statistical systems, their institutions and people, has become more urgent and critical in ensuring that progress is effectively monitored and gaps timely addressed. It is important to note that the SDGs include statistics development as a vital outcome and rightly so. It is no assurance however that governments will significantly give priority to statistics improvement given many competing concerns especially in many developing economies where appreciation for statistics, let alone investment, has been historically wanting. The Country Report on Support to Statistics (CRESS)

The MDGs provided impetus and brought about many opportunities for improving statistics and developing capacities in the national statistical systems. The monitoring of the MDGs may have increased some degree of appreciation for statistics and considerable investment in statistics development. The MDG Indicators framework proved useful not only for measurement and monitoring purposes but also as a reference for capacity building and statistical advocacy, including advocacy for increasing investments in statistics. The SDGs, being a more comprehensive, wide ranging framework, will require considerably bigger financial investment and extensive and better human capital to improve the quality of existing statistics and to produce new ones. Many national statistical systems, especially in developing economies and low-income countries may be pushed to the limits of their capacity and resources to address these bigger expectations. Development partners, whose resources are admittedly also not limitless, will need to work more effectively together with national statistical systems in building data ecosystems to support the SDGs as well as national development goals. Increased investment that is based on a good strategic plan is key to realizing outcomes in statistics development. This principle provides a premise for the CRESS. Cognizant that financing is essential to the collection of statistics to support national and global development needs, PARIS21 conceptualized and introduced the CRESS approach to gathering information and documenting the financing of statistics at the country level. PARIS21 aims to build on their experience with the PRESS which is perhaps the most comprehensive if not the only source of information on partnership investments in statistics currently available. In a nutshell, the CRESS aims to inform stakeholders of the patterns and dynamics of financing statistics at the country level and how such information may be used in advocacy towards increased investment in statistics, in expenditure planning for statistics, and in monitoring and evaluation of statistics development, including financing and investment. More specifically, it seeks to provide information on the financing of the NSS by domestic and external

About PARIS21

Since its establishment in 1999, the Partnership in Statistics for Development in the 21st Century (PARIS21) has supported the strengthening of statistical capacity of developing countries to improve their provision, availability and use of high quality data; foster dialogue within societies on the use of data for accountability and evidence-based policy making; and promote the well-being of people through contributing to the achievement of national and international development goals. It remains a catalyst for promoting evidence-based standards and policies among countries, bringing together producers and users of statistics in a shared goal towards statistical development and better development outcomes.

3

sources, the priority areas of investment, and future financing requirements or plan; validate and monitor the extent of financing for statistics by development partners; and understand development partners’ focus of investment in statistics and possible areas for harmonization.

The CRESS initiative started in 2011 using as pilot cases the national statistical offices of Cameroon, Ethiopia, Malawi, and Senegal in the Africa region with support from PARIS21. CRESS was introduced in the Asia-Pacific region in 2015 with the Philippines and Vietnam as the first countries to undertake the exercise. 2. Why CRESS in Cambodia?

The Statistics Law, amended in 2015, and the 2008-2015 Statistics Master Plan (SMP) of Cambodia have provided impetus for reforms in Cambodia’s national statistical system. With concrete legal basis and a strategic plan, the environment seemed set to pave the way for improved capacity and increased resources for statistics, especially the National Institute of Statistics (NIS). Through the Statistics Law and with strong political support of the Royal Government of Cambodia (RGC), the central role of NIS in the management and coordination of the Cambodia statistical system has finally earned a long sought and much deserved recognition. The Law has clarified and delineated statistical responsibilities among the various ministries and agencies stressing increased ownership and accountability. It has also established high-level governance mechanisms for a more coordinated and integrated statistical system. A ministerial level statistics advisory council is in place to provide strategic directions and policies and is supported by a statistics coordinating committee that is tasked to translate policies into integrated and coordinated action plans and monitor their implementation.

Figure 1. Organizational framework of the Cambodia national statistical system1

1 Roadmap on the formulation of the National Strategy for the Development of Statistics in Cambodia. PARIS21 and NIS. February 2016.

4

Since 2008, Cambodia has conducted several major statistical activities like the 2008 Census of Population, labor force surveys, an intercensal population survey, annual socio-economic surveys, demographic and health surveys, and more recently, the first ever Economic Census, conducted in 2011, and the first Agricultural Census, completed in 2013, with support of both the government and international development community. Data dissemination has significantly improved. More than ever, a lot of data about Cambodia has been generated for users. There have been significant improvements in the statistical system. For instance, salaries for statisticians have generally increased albeit still far from competitive levels. Many statistical personnel have had greater access to a wide range of international statistical training and workshops. Newer technology, both in knowhow and equipment, for improved statistical operations has been acquired. Coordination and cooperation among data producers have improved. International support for statistics development has been steady and significant because the Cambodia national statistical system has had a considerably long history of partnerships with international development partners and donor agencies.2 For some time, in particular since 2005 at the start of implementation of the Statistics Law and Statistics Master Plan, government funding and its share of total financing for statistics have been increasing. Recent data however indicates a picture far from rosy. The 2012 Mid-Term Review of the SMP revealed that almost 91 percent of the total budget for statistics in Cambodia was allocated to the NIS in 2011 while the rest was shared among the ministries and agencies with designated official statistics.3 The same review mentioned that while government’s budget allocation for NIS increased dramatically during the period 2008-2011, external financing remained disproportionately high. Donor contribution was said to have dropped from 87% of total expenditure for statistics in 2008 to 65% in 2011. With the next Census of Population coming up in 2018 and other major annual and periodic surveys expected to be conducted in the period 2017-2019 the declining trend in external statistical financing is putting pressure on the government to look for more local resources for statistics. How the national statistical system will cope with this has become a key concern. This provided an opportune basis for the preparation of the CRESS in Cambodia aimed at providing more informed analysis of the situation of the financing of statistics in the country towards better decisions and concrete action by the appropriate stakeholders. The preparation of the CRESS Cambodia is borne of a series of progressive partnerships between PARIS21 and NIS that include high level discussions towards the formulation of the statistics law, capacity building on national strategy for the development of statistics (NSDS), and the preparation of a roadmap for the formulation of a NSDS. With the NSDS roadmap set to be implemented soon, the CRESS Cambodia is seen as a timely and relevant initiative to provide concrete basis for the preparation of a medium-term financing plan for the

2 Roadmap for the National Strategy for the Development of Statistics of Cambodia. PARIS21 and NIS Cambodia. February 2016. 3 Refers to all statistics covered by the Sub-decree on Designated Official Statistics issued by the Prime Minister of the Royal Government of Cambodia emanating from the Statistics Law of 2005 and subsequent amendments and updates; refers to all basic statistics, administrative data, and sectoral statistics, including those generated through censuses and surveys, by the NIS and designated statistical units in selected ministries and agencies.

5

NSDS. The timing is crucial as the national statistical system braces for a number of emerging challenges such as the

• the adoption of the SDGs and the need to monitor them; • the updating of the National Strategic Development Plan and expectations for the production of

timely and accessible quality statistical information for its monitoring; and • the continuing evolution of the national statistical system.

The CRESS Cambodia aims in particular to provide information and basis for NIS to explore solutions for separate or dedicated budget allocation for statistics within the framework of the government’s new policy on program budgeting. The CRESS Cambodia seeks to present stakeholders in both government and international development community with concrete evidence of how statistics is financed in terms of levels and trends as well as the gaps and challenges in budget allocation for statistics in the country. 3. The CRESS Cambodia

3.1 Methodology

The methodology for CRESS Cambodia was adapted from earlier initiatives, especially the CRESS pilot study in the Philippines in 2015. The methodology as adapted is described in the CRESS Cambodia Concept Note in Annex A.

3.1.1 Organizing the CRESS Cambodia Country Team As a country-led exercise, it was deemed important that the key CRESS activities are guided, managed and undertaken by the appropriate institution in the country. The NIS, as the primary stakeholder, formed a CRESS Cambodia Country Team composed of key NIS official and selected staff members, supported by national and international consultants engaged by PARIS21 for the project to manage and coordinate the process. The Team helped develop and adapt the CRESS methodology and tools for Cambodia, managed the distribution of questionnaires and collection of responses, validated the responses and encoded the data into the CRESS database with guidance and assistance from PARIS21 and international consultant. The existing Statistics Coordinating Committee (SCC), chaired by the NIS Director General, was initially considered to serve as steering committee to provide direction, resolve issues, and review the findings and recommendations. 3.1.2 Launching forum As in many other countries, the Cambodia national statistical system involves multiple stakeholders comprising institutions and individuals that collect, manage, disseminate, and use statistics for policy, plan, and program formulation, implementation, and monitoring and evaluation. Effectively, these stakeholders are the primary sources of information for the CRESS.

6

The CRESS Cambodia was officially launched during a forum held on 23 September 2016 in Phnom Penh Cambodia with 54 participants coming from NIS and ministries and agencies with designated statistics based on the Statistics Law and selected international development partners (see programme, list of participants, and presentation materials in Annex B). The international consultant presented the concept and methodology of the CRESS as well as snapshots of the items of information to be requested. In addition to orienting the participants about the CRESS exercise and clarifying expectations, the forum was also aimed to advocate support for the CRESS. 3.1.3 Scope of the data

The CRESS Cambodia theoretically covers all types of institutional resources, statistical activities, and details of financing i.e., major expense categories.

a. Institutional resources are either

(1) national resources, which are financial resources used/intended to support the

organizational requirements and statistical operations of the national (government) statistical system ― the national statistical office or authority, organic or designated statistical units of government ministries and agencies that produce designated statistics, other government agencies that collect and report official statistics, and state institutions that provide education and training on statistics; or

(2) external resources, which refer to grants, loans, or aid provided by development partners or donor agencies, private institutions, and nongovernment organizations as targeted support for statistical developmental activities of the national statistical system through a specific or group of government ministries/agencies/institutions. It should be noted that some of external resources support standalone statistical development programs or projects while some support statistical activities that are components of a bigger development project/program, e.g., monitoring and evaluation, research, etc.

b. Statistical activities include but are not limited to:

(1) regular and periodic censuses and surveys, (2) administrative data collection, (3) development, updating, and compilation of statistical/data frameworks and indicators, (4) data processing and quality assurance, (5) data analysis, dissemination and communication, (6) development, improvement, and updating of information systems and databases, (7) development/improvement of statistical methodologies and data collection/capture

instruments, (8) statistical education and training, and other institutional and human resource capacity

development activities, (9) policy and standards formulation, planning, and coordination, and (10) monitoring and evaluation.

7

The UNECE Classification of Statistical Activities (Annex C) was used as basic reference for identifying the statistical domains of activities financed or to be financed. The CRESS Cambodia made a few minor modifications to adapt relevant developments in the country.

c. Details by general expenditure category for regular or recurring statistical activities as well as

periodic or multi-year statistical activities and special projects were requested as follows: (a) personal services, which include salaries and wages, and other personnel emoluments of

statistical personnel; (b) maintenance and other operating expenses of statistical agencies/institutions and units; and (c) capital outlay or investments for physical and technological infrastructure, systems, and

equipment, e.g., construction of building for office and data operations (such as processing and dissemination centres), acquisition of computers, electronic devices, and software, data/information system and application development or upgrading, connectivity, etc..

3.1.4 Temporal coverage The CRESS Cambodia covers financing for statistics during the period 2013-2016 (past) and planned expenditure or financing commitment for the period 2017-2019 (future). 3.1.5 Data sources

The target respondents for the CRESS Cambodia included the NIS, twenty-seven (27) ministries and agencies with designated official statistics based on the amended Statistics Law 2015 and the Ministry of Planning (MOP) for government financing, and twenty (20) bilateral and multilateral development partners that have supported statistics in Cambodia at one time or another. The study did not consider academic, research, and training institutions, local government units, non-government organizations, and the private sector. There was prior information that no academic or training institution in Cambodia offers statistics education or training other than the Center for Planning and Statistics of the Department of Personnel of the Ministry of Planning. The Country Team also agreed to exclude the non-government organizations and private sector as there was no known instance of any private sector financing for statistics in the past and the LGUs solely for manageability and time constraint consideration in covering a significant number of local offices.

Twenty (20) of the 29 target ministries and agencies returned filled out questionnaires and were accepted as actual respondents to the study from the government side, while seven (7) of the 15 target organizations constitute the actual respondents from the development partners’ side. The full list of target and actual respondents and other respondents’ information is in Appendix D.

Supplemental and supporting data and information were gathered from the following documents:

• the Budget Law of the Royal Government of Cambodia, specifically the matrix of state

expenditures by ministry from 2013-2017 (original version in Khmer); • Annual reports of the Cambodia statistical system as published by NIS, 2013-2014; • Websites of selected ministries and agencies and development partners; and • PRESS reports for 2013, 2015, and 2016, PARIS21 website.

8

3.1.6 The instruments

CRESS Cambodia adapted the Philippines’ questionnaires with some adjustments for better clarity and to collect additional information. Whereas CRESS Philippines used 5 questionnaires, 3 for government and 2 for development partners, CRESS Cambodia added an additional questionnaire for government.

The following questionnaires were distributed to the 29 target ministries and agencies:

• Questionnaire 1 asks for basic respondent information and summary information on financing

levels and sources, number of statistics projects, and statistical positions. • Questionnaire 2 asks for details of financing from domestic sources for 2013-2016 including

regular funding and special or nonrecurring statistics projects. • Questionnaire 3 asks for the profile of statistics projects financed with external assistance during

the period 2013-2016. • Questionnaire 4 asks for the profile of planned statistics projects with proposed or committed

financing for 2017-2019.

The following questionnaires were designed for and distributed to 15 target development partners:

• Questionnaire 5 requests for the profile of statistics projects financed by development partners during the period 2013-2016.

• Questionnaire 6 requests for the profile of planned statistics projects to be financed by development partners in 2017-2019.

The CRESS Cambodia questionnaires are in Annex D.

As agreed with the Country team, questionnaires 1-4 were translated into Khmer by the national consultant to address language concerns and ensure clarity and ease of understanding of the questions and instructions by the respondents as well as to promote accurate responses. The Khmer version was reviewed and tested by the Country Team to make sure that the translation accurately reflects the original questions. The questionnaires both English and Khmer versions were originally planned to be pilot tested by NIS and MOP before finalization and distribution to the rest of the respondents. However due to time constraint and initial confusion as to the focal points, the pilot test was dropped.

3.1.7 Data collection The printed questionnaires together with the metadata, definitions, instructions, and a modified UNECE CSA were distributed to all respondents through an official letter of request from the NIS Director General from September until as late as December 2016. The distribution of questionnaires had to be extended several times to consider the series of official and public holidays in Cambodia.

The Country Team followed up the return of accomplished questionnaires through phone and email communication providing additional explanation as and when needed. The Country Team also conducted visits in some ministries and agencies to follow up, clarify issues, and guide respondents

9

in accomplishing the questionnaires. During the two (2) country missions conducted by the international consultant, one with the PARIS21 Secretariat, several ministries and agencies were additionally visited to further encourage submission of questionnaires. Special orientation meetings were held for selected ministries and agencies to provide more guidance and advice in answering the questions and in using estimation techniques for certain cases. Most of the accomplished questionnaires were submitted in MS Excel format through email. Only a few returned the questionnaires in print. Some ministries and agencies submitted the Khmer version of the questionnaires necessitating translation back into English by the national consultant. 3.1.8 Data validation The Country Team reviewed, validated, and encoded the data as soon as the filled-out questionnaires were received. A designated database administrator from the team managed the initial data review and encoding into the CRESS database. The updated database was shared with the other team members and the consultants as soon as new data came. The CRESS Cambodia database was built after the Philippine’s format but with some adjustments to consider differences in the questionnaires. The database consists of 6 components each corresponding to one of the 6 questionnaires used in the study. Each data item in each questionnaire was assigned a database column. Analytical tables were generated from the database to facilitate analysis of results and formulation of findings. Further data validation by the national and international consultants was done to make sure that all responses were clarified and accounted for.

In addition to data from the CRESS questionnaires, the CRESS Country Team also used data and information from the websites and monitoring plans and reports of the ministries, agencies and development partners as well as from the Royal Government’s Budget Law documents. As most of the documents were in Khmer, translation was necessary and was provided by the national consultant.

Throughout the duration of the data collection and validation exercise, the CRESS Country Team and the international consultant documented the observations and challenges as well as the lessons learned.

10

4. The CRESS Cambodia results and findings Results presented in this section are subject to the limitations of the study and the responses.

4.1 Analysis of support to statistics Total financing for statistics as reported by twenty (20) government ministries and agencies amounted to US$ 35.8 million for the four-year period 2013-2016. About 73 percent of funding came from the government while 27 percent was received from external sources (Figure 2). The responding ministries and agencies reported a total of US$ 26.3 million that were allocated to statistical activities as part of their annual budget and US$ 9.5 million that were provided by international development partners for statistics projects implemented by government over the same period (RG Table 1.2).

Figure 2 Figure 3

About the data

• Data used in the following analysis is based on responses from 20 of 29 target ministries and agencies from the government side, including NIS, and 7 of 15 target international development partners.

• Of the 20 government respondents, 16 ministries/agencies provided at least 1 data point for the period 2013-2016. Only 13 submitted reasonably complete data for the same period. On the other hand, 15 ministries/agencies provided data for 2017-2019. There are some items (questions or years) with no response/s. Because of this, some aggregations may only represent the minimum levels of financing.

• Of the 7 responding development partners, technically only 4 returned accomplished questionnaires; 3 provided reference documents from which data was extracted by the consultants and transferred to the questionnaires.

• Data does not include financing by private or nor government organizations, local government units, and academic and research institutions.

11

Financing for statistics appears to have been on the decline during the period, both from the government and international development partners. Data shows that total funding for statistics continuously dropped from US$ 12.9 million in 2013 to US$ 9.3 million in 2015 and rose again to US$ 7.8 million in 2016 (Figure 3, RG Table 1.1). The general downtrend is mainly attributed to the decrease in external funding which usually comes with the conduct of major statistical activities in the country as follows:

2013: Census of Agriculture, Cambodia Intercensal Population Survey, and (annual) Cambodia Socio-Economic Survey;

2014: Cambodia Demographic and Health Survey, Cambodia Industrial and Economic Survey, and Cambodia Socio-Economic Survey;

2015: Cambodia Socio-Economic Survey; and 2016: Cambodia Socio-Economic Survey.

The share of external funding declined from 43.4 percent in 2013 to 13.2 percent and 13.5 percent in 2015 and 2016, respectively. Consequently, the share of government funding significantly increased from 56.6 percent in 2013 to 86.8 percent in 2015, although slightly decreasing to 86.5 percent in 2016. (Figure 3, RG Table 1.2) Data from the international development partners, that is, seven (7) of the fifteen (15) bilateral and multilateral agencies requested to take part in the survey, shows a total of US$ 21.3 million worth financial and technical assistance for statistics provided to selected government ministries and agencies during the period 2013-2016 (DP Table 1.1). The amount registered was some US$ 11.8 million more than what was reported by the responding government ministries and agencies. Based on the consolidated PRESS reports for 2015 and 2016 4, the total financing for statistics in Cambodia from development partners for the period 2013-2016 amounted to at least US$ 25.1 million (with undercoverage and with the exclusion of regional programs). (DP Table 13)

4.1.1 Support from domestic (government) sources

• The total government allocation for statistics in the 20 responding ministries and agencies decreased from US$ 7.3 million in 2013 to US$ 5 million in 2015 then increased to US$ 6.8 million in 2016 (Figure 4, RG Table 2.1). The increase from 2015 to 2016 however was primarily due to the unusually high budget allocation for statistics reported by the Ministry of Health in 2016 amounting to US$ 3.1 million, which is said to be documented in Cambodia’s Third Health Strategic Plan 2016–2020, compared to a much smaller budget in 2015; no allocation was reported in 2013-2014. In addition, a few other ministries reported data only for 2015 and/or 2016.

4 Partnership Report on Support to Statistics. PARIS21 website.

12

Figure 4

• Based on the reported allocation of ministries and agencies with complete data for the period

2013-2016 however, the total budget for statistics may have in fact further declined, and quite significantly, in 2016 (Figure 4, RG Table 2.2) due to fewer major statistics projects implemented in 2015-2016.

• The bulk of government allocation for statistics went to the NIS with 73.7 percent followed by the Ministry of Health (MOH) with 12 percent, and the Ministry of Education, Youth, and Sport (MOEYS) with 4.9 percent (RG Table 2.1).

• Data shows a steep decline in government allocation to NIS from 2013 to 2016 (Figure 5). In contrast, data from a separate budget monitoring report compiled by NIS reveals a continuously increasing trend in government allocation to NIS during the period 2005-2013 covering the implementation of the Statistics Law and the 2008-2015 SMP. The same report indicates a general decline in external support to NIS which was mainly intended for the conduct of major national censuses and surveys.

Figure 5

13

• The total regular budget allocation for statistics was overall on a downtrend during the period 2013-2016. Slightly increasing from US$ 5.7 million in 2013 to US$ 6.1 million in 2014, the total regular budget allocation for statistics dropped to US$ 4.5 million in 2015 and US$ 2.9 million in 2016 (Figure 6, RG Table 4.1). Data shows that while the budget for NIS showed a declining trend, allocation for statistics of the remaining responding ministries and agencies gradually increased during the period. Upon closer look, however, the upsurge was in fact driven only by the sharp increase in the budget of the MOEYS in 2016 mainly for the major improvement of the educational management information system.

Figure 6

• Regular budget allocation for statistics constitutes 73.3 percent (equivalent to US$ 19.3 million),

of the total government allocation for statistics for the period 2013-2016 while budget allocation for special and nonrecurring statistics projects, including government counterpart funding, comprises 26.7 percent (or US$ 7 million) (RG Table 4.2). On an annual basis, the share of regular allocation to total budget for statistics was increasing from 78.4 percent in 2013, 85.2 percent in 2014, and 89.3 percent in 2015 but reduced to only 43.5 percent in 2016 (Figure 7, RG Table 4.2) due to the large special budget reported by the Ministry of Health for 2016.

Figure 7

14

• Only five (5) of the 20 responding ministries and agencies were allocated budget for special and

nonrecurring statistics projects with NIS getting 50.4 percent (equivalent to US$ 3.5 million) of the total government special funding for statistics (RG Table 4.2). The other ministries and agencies with special allocation for statistics were the MOH, MoEYS, the Ministry of Social Affairs, Veteran and Youth Rehabilitation (MOSVY), and the Ministry of Culture and Arts.

• On average, government budget allocation for statistics as percentage of the total state

expenditure of the 14 responding ministries and agencies with complete data for the period 2013-2016 (including NIS as part of MOP5) was less than two percent. More specifically, the share of government allocation for statistics in ministry/agency expenditure decreased from 1.6 percent in 2013 to less than one percent in 2014 and 2015, to a very low 0.2 percent in 2016 indicating a diminishing priority accorded to statistics in the ministries and agencies (Figure 8, RG Table 10.1).

• As percentage of total government expenditure, government allocation for statistics of the same ministries and agencies was even much lower, just hovering above zero during the period 2013-2016 (Figure 8, RG Table 10.2).

Figure 8

• The total government budget for statistics of the responding ministries and agencies was on

average less than one tenth of one percent of the gross domestic product at current prices (RG Table 10.3).

• The 20 responding ministries/agencies reported a total of 21 statistics projects financed by either or both government and development partners in 2013-2016 and 13 projects to be allocated budget in 2017-2019. Seventeen (17) of the 21 statistical projects funded in 2013-2016 were either fully supported by an international development partner or co-financed by government and international development partner/s or among the latter. (RG Table 6)

• It is noted in particular that while many of the statistics projects financed in 2013-2016 were

implemented by NIS, only one statistics project was identified by NIS for financing in 2017-2019. 5 NIS is a directorate general under the Ministry of Planning.

15

Based on the designated official statistics schedule, NIS is expected to conduct at least the decadal population census in 2018, quinquennial demographic and health survey in 2018 or 2019, and annual socio-economic surveys.

• On average, during the period 2013-2016, 74.4 percent of authorized statistics positions in the

20 responding ministries and agencies were in the NIS while 25.6 percent were with the designated statistical units of the ministries and agencies. The total number of authorized positions did not vary much from year to year during the period. In 9 ministries and agencies for which data for both authorized and filled positions are available, the number of statistics positions remained constant in 2013-2016. Most of the positions were filled up to 98 percent throughout the period. (Figure 9, RG Table 11.1)

Figure 9

• Six (6) ministries did not indicate any statistical position or personnel while four (4)

ministries/agencies could only provide data for some years because of (a) difficulty in getting the data, (b) the statistical unit and positions were only established in the latter years during the period 2013-2016, and (c) many of the staff are doing multiple functions such as planning and updating of administrative data. One agency declared a lone statistics position but reported a significant number of personnel doing statistical functions.

• Data on the total number of personnel in the ministries/agencies could not be made available hence no information on the proportion of statistical human resources to the total government workforce could be provided.

• For purposes of the CRESS exercise, statistics positions (or personnel) would include those

pertaining to statisticians, IT personnel, economists, analysts, enumerators, data encoders, and administrative assistants, provided that their job entails database management, data gathering, information system maintenance, and other statistics-related activities such as data collection, analysis, interpretation, presentation, and organization of numerical and non-numerical data. Also, considered as statistics personnel are those personnel that are assigned/working in dedicated statistics units which primary or full functions comprise the production, management, coordination, analysis, and/or dissemination of statistics, as well as education and training in statistics. Non-technical positions in statistical units exclusively performing statistical functions are also included.

16

4.1.2 Support from external sources

As reported by the government

• Only five (5) government ministries and agencies reported having received assistance for statistics from development partners, namely, NIS, MAFF, MoEYS, Ministry of Women Affairs, and MOP. Total external support for statistics had a cumulative value of US$ 9.5 million for the period 2013-2016 constituting 26.7 percent of cumulative total financing for statistics during the same period (RG Table 12). NIS benefitted the most from external assistance receiving US$ 8.3 million, or 87.3 percent of total external financing support, for the conduct of several major national statistical activities during the period (Figure 10, RG Table 13.2).

Figure 10

• On an annual basis, however, the share of external support to total financing had dropped

dramatically from 43.3 percent in 2013 to 22.8 percent in 2014, to 13.2 percent and 13.5 percent in 2015 and 2016, respectively, which could also be attributed to the unusually high government allocation for the MOEYS and MOH in 2016 (Figure 3, RG Table 1.2).

• Sweden/Sida was the top provider of financial and technical assistance during the period with at

least 28.6 percent of the total external financing support or a total of US$ 2.7 million (Sida co-financed another statistics project with UNESCO) (Table 1, RG Table 13.3). Other development partners that were reported to have provided significant support for statistics were Australia/DFAT, USAID, and UNFPA.

17

Table 1 Statistics projects and external financing, by source: 2013-2016

Source Number of statistics

projects 1/

Financing support (US$)

Share of external

financing (%)

ADB 1 15,000 0.2

Australia/DFAT 1 1,000,000 10.5

EU 1 184,567 1.9

FAO 1 376,749 3.9

Japan/JICA 4 393,975 4.1

Korea/KOICA 1 61,377 0.6

Sweden/Sida 4 2,729,356 28.6

UNFPA 1 709,792 7.4

UNFPA and Australia/DFAT 1 1,052,031 11.0

UNICEF 1 126,805 1.3

UNICEF/Sweden/Sida, UNESCO 1 700,219 7.3

UNSD/DFID 1 46,800 0.5

USAID 1 1,500,000 15.7

WB 1 350,000 3.7

WHO and UN Women 1 200,000 2.1

Unspecified source 2 98,310 1.0

Total 23 9,544,981

1/ counted per project that was provided assistance; may refer to same project/s that were co-financed with other development partner

• International development partners supported 15 projects primarily for collecting data or

improving economic and demographic and social statistics (based the UNECE Classification of Activities) (Figure 11, RG Table 14.1).

Figure 11

18

• Projects financed in 2013-2016 included the first agriculture census in Cambodia, an intercensal population survey, socio-economic surveys, an industrial and economic survey, demographic and health survey, strengthening statistics on education and gender, and improving capacity in poverty analysis (Table 2, RG Table 7).

Table 2

Statistics projects with financing support from development partners as reported by government: 2013-2016

Statistics project Reference year/year

implemented

Financing support (US$)

Recipient ministry/agency

Development partner/s

1 Cambodia Intercensal Population Survey 2013 825,429 NIS Japan/JICA, UNFPA

2 Cambodia Socioeconomic Survey 2013 679,025 NIS Sweden/Sida 3 Census of Agriculture 2013 3,876,749 NIS FAO, Australia/DFAT,

Sweden/Sida, USAID

4 Improving the collation, availability and dissemination of national development indicators, including MDGs

2013 46,800 NIS UNSD/DFID

5 Cambodia Industrial and Economic Survey

2014 211,503 NIS Japan/JICA

6 Cambodia Demographic and Health Survey

2014 1,277,048 NIS UNFPA, Australia/DFAT,

Japan/JICA, Korea/KOICA

7 Cambodia Socioeconomic Survey 2014 532,452 NIS Sweden/Sida 8 Statistical Literacy and M&E capacity

development in support to NSDP/CMDG and Sector M&E System Consultant

2014 10,200 MOP WB

9 Cambodia Socioeconomic Survey 2015 172,361 NIS Sweden/Sida

10 Unspecified project on education statistics

2015 88,110 MoEYS Not indicated

11 Enhancement of Poverty Analysis Capacity Project

2015 350,000 NIS WB

12 Statistical Business Registers 2015 15,000 NIS ADB

13 Gender statistics 2013-2015 30,000 MoWA Japan/JICA

14 National Survey on Women's Health and Life Experiences in Cambodia

2014-2015 200,000 MoWA WHO and UN Women

15 Cambodian Intercensal Agricultural Survey

2016 184,567 MAFF EU

16 Cambodia Socioeconomic Survey 2016 345,518 NIS Sweden/Sida

17 Strengthening Education statistics 2013-2016 700,219 MoEYS UNICEF, Sweden/Sida, UNESCO

Total cost

9,544,981

19

As reported by development partners

• The total financing commitment by the seven (7) responding development partners in 2013-2016 amounted to US$ 21.3 million (DP Table 1), more than double the total amount reported by the government or a discrepancy of about US$ 11.8 million (Figure 12). Based on reported data from both sides, there is basis to believe that (a) respondents (officers that accomplished the questionnaires) were not aware of or did not have access to information on external financing support received by responding units in the ministries and agencies; and (b) respondents had little coordination with their counterpart in the development partners.

• Using data reported by international development partners, the total financing for statistics in Cambodia would reach US$ 47.6 million instead of the US$ 35.8 million reported by government (Figure 12). The share of external funding for the period 2013-2016 would likewise increase to 44.8 percent from 26.7 percent based on government data.

Figure 12

• A total of 36 statistics projects were supported by development partners during the period

2013-2016, 19 projects more than what was reported by the government (DP Table 3). While all but one of the recipient ministries/agencies submitted accomplished questionnaires, most did not report many of the projects cited by the responding development partners. As mentioned earlier in the report, there is indication of weak coordination between ministries and agencies and development partner or inaccessibility of information about financing and technical cooperation.

• Fourteen (14) of the 36 projects were solely financed by development partners and 7 projects were co-financed with other development partners while information was not available for the remaining 16 projects (DP Table 4). Fifteen (15) of the 36 projects financed in 2013-2016 were a component of bigger projects, 12 were standalone projects, and 9 projects were not classified (DP Table 4).

• The biggest source of external funding was Australia/DFAT providing 41.9 percent of total financing support to statistical projects followed by Sweden/Sida with 18.1 percent, WHO which contributed 17.3 percent, and UNFPA with 11.5 percent (Figures 13 & 14, DP Table 1).

20

Leading sources of external financing for statistics, 2013-2016

Figure 13

Figure 14

• It must be noted however that Sweden/Sida was not the sole source of funding for one project,

the technical assistance to NIS, and UNFPA was not the lone source for two projects, the CDHS and CIES which were both conducted in 2014. For these projects, Sweden/Sida and UNFPA served as lead development partner and coordinator of the co-financing arrangement with government.

• In terms of original or ultimate source of financing, Australia/DFAT remains the top provider of

external financing for statistics contributing 41.9 percent of total funding in 2013-2016, followed by WHO with 17.3 percent, and Sweden/Sida with 17.1 percent. The other sources of external resources which include development partners that did not respond to the survey are listed in the table below (Table 3, DP Table 5).

21

Table 3 External financing for statistics as reported

by development partners by ultimate funding source

Ultimate funding source 1/

Financing support (US$)

Share to total financing (%)

Australia/DFAT 2/ 8,938,433 41.9 EU 207,662 1.0 Japan/JICA 198,993 0.9 Korea/KOICA 61,377 0.3 Sweden/Sida 2/ 3,639,905 17.1

UNDP 2/ 1,148,333 5.4

UNFPA 2/ 1,156,016 5.4

UNICEF 2/ 1,350,000 6.3

USAID 857,876 4.0 WB 2/ 17,395 0.1 WHO 2/ 3,694,658 17.3 HSSP2 3/ 40,000 0.2

Total 21,310,648

1/ Refers to the original source of funding; may or may not be a responding development partner.

2/ Refers to responding development partner.

3/ Health Sector Support Programme, a multi-donor collaboration among six development partners, namely, the AFD, DFID, AusAID (now DFAT), WB, UNICEF, and UNFPA, supporting the Royal Government of Cambodia.

• The biggest recipient of external financing support from international development partners during the period 2013-2016 was the NIS getting 37 percent of all funding received (Figure 15, DP Table 2). The other top recipient agencies were the MOP and MOH. About 10 percent of financing went to unspecified recipient ministries/agencies.

Figure 15

22

• Twenty (22) of 23 statistics projects with information on the nature of support were a combination of financial and technical assistance (DP Table 9).

• Most of the projects funded by the international development partners were on demographic and social statistics and environmental and multi-domain statistics, in particular, statistics on gender and special population groups, and health (DP Table 6.1 & 6.2, Table 4). These include a number of major national statistical activities such as the first agriculture census, the quinquennial demographic and health survey, intercensal population survey, industrial and economic survey, and the annual socio-economic surveys which were all conducted by NIS. The other projects dealt with basic data collection and analysis in support of bigger projects or specific sectoral or multi-domain subjects being advocated by individual development partners.

Table 4 Number of statistics projects with financing support

from development partners by statistical domain

Statistical domain 1/ Number of statistics projects 2/

Demographic and social statistics 23

Economic statistics 4

Environment and multi-domain statistics 24

Methodology of data collection, processing, dissemination and analysis 6

Strategic and managerial issues of official statistics 4

Total number of statistics projects 36 1/ based on a modified UNECE Classification of Statistical Activities for CRESS Cambodia 2/ Some statistics projects are classified under one or more statistical domain/s and are counted under each domain or subdomain; do not add up to total.

• The weighted average duration of projects during the period 2013-2016 supported by the

responding international development partners was 40 months (DP Table 10).

4.1.3 Financing prospects for 2015-2017 As reported by the government

• Fifteen (15) of the 20 responding ministries and agencies, including NIS, provided information on planned or proposed allocation for statistics in 2017 amounting to US$ 22.1 million. However, only half of these ministries and agencies, excluding NIS, provided indicative budget for 2018 and 2019, which would cost US$ 3.8 million and US$ 3.9 million, respectively (RG Table 15.1).

23

• Ninety (90) percent of the total proposed financing in 2017 and all of the indicative budget for 2018 and 2019 are expected to come from the government (RG Table 15.2).

As reported by the development partners

• Meanwhile, six (6) of the 7 international development partners, namely, Australia/DFAT, Sweden/Sida, UNDP, UNFPA, WB, and WHO, have indicated continued support for statistics in 2017-2019 through 7 projects with a total cost of US$ 8 million (DP Table 11). The total external financing for statistics in 2017-2019 is less than half of the support provided in 2013-2016 as planning and coordination are still underway for the financing of future statistical activities.

Figure 16

• A substantial share of the total prospective financing amounting to US$ 5 million or 62.7 percent

is to come from Australia/DFAT for a project that aims to collect and analyze data for identifying poor households in Cambodia. MOP is implementing the said project and another one with NIS to be funded by WB on capacity development to produce and analyze poverty and inequality indicators and to improve price data collection. MOP thus stands to receive a considerable share of at least 62.7 percent of the total financing assistance for statistics in 2017-2019. (DP Table 12)

• The other government institutions that can expect continued external support are MOH for its health information system and immunization survey which is planned for funding by WHO and NIS for the conduct of the 2018 General Census of Population which is to be supported by UNFPA (DP Table 12).

• The statistics projects that will have financing assistance from development partners in 2017-2019 are as follows:

24

Table 5 Statistics projects for financing support by development partners: 2017-2019

Statistics project Financing cost (US$)

Share to total cost

Development partner

Recipient ministry/agency

1 Identification of Poor Households Program Phase 3 5,038,000 62.7 Australia/DFAT MOP

2 National Statistics 1,735,577 21.6 Sweden/Sida and EU 1/ NIS

3 Partnership for Gender Equity III 60,000 0.7 UNDP Unspecified

4 2019 General Population Census of Cambodia 242,963 3.0 UNFPA NIS

5 Enhancement of poverty analysis capacity project 332,606 4.1 WB MOP/NIS

6 Health system information and evidence 315,070 3.9 WHO MOH

7 National immunization programme (NIP) 310,200 3.9 WHO MOH

Total cost 8,034,416

1/ Based on project document; subject to changes/revisions. Data used in the above analysis can be found in Appendix C. 4.2 Limitations The study was planned to be conducted within a relatively short timetable which may have contributed to some limitations in terms of coverage and completeness. The more critical issues have to do with underestimation of financing of statistics both from domestic and external sources. While it has been clarified at the onset that the study was to be limited to the ministries and agencies with designated official statistics and that it would not cover support from the private sector, non-government organizations, academic and research institutions, and local government units, there are other sources of underestimation which are:

• non-response; non-submission of questionnaires by 9 ministries and agencies with designated official statistics and some international development partners that are known to have provided support for statistics in the past such as Japan/JICA and the Food and Agriculture Organization;

• incomplete data (or item non-response) from some responding ministries and agencies and development partners;

• non-reporting or underreporting of financing of key statistical frameworks and indicators related to administrative-based data systems such as the generation of foreign trade and investment statistics, vital registration, and environment-related statistics; data management and dissemination; statistical management, policy, coordination, and advocacy activities; and statistical capacity building initiatives;

• confidentiality considerations in the provision of full information by one agency;

25

• non-inclusion of country support either as standalone or part of global, regional, or multi-country programs of the UN Statistics Division, UNESCAP, UNSIAP, ASEAN Secretariat, IMF, ADB, etc.;

• difficulty in identification and classification of statistical activities, operations, human resources, and capital outlays from the budget of most ministries and agencies; and

• inability of ministries and agencies in indicating financial commitment for future statistical activities.

4.3 Challenges While the CRESS Cambodia initiative benefitted a lot from the Philippines’ pilot experience, there were issues and challenges throughout the process that can only be attributed to the country’s situation. It is important to document and discuss these challenges so that these can be addressed for similar future undertaking.

• Lack of priority, focus, and capacity in managing and undertaking the CRESS process

Although NIS designated a key official and selected staff to the CRESS Cambodia country team, the demand of many competing priorities at least during the implementation period had limited the capacity of the concerned official and staff in coordinating and carrying out the necessary activities. Of the initial 6 staff designated to constitute the country team, only 2 technical staff were left to undertake multiple tasks and which compelled the international and national consultants hired by PARIS21 to assist NIS to cover more areas of work than originally required based on their terms of reference. It took a while before the questionnaires were finalized largely because of little feedback on the changes and improvements made. It took longer to distribute the questionnaires because of certain protocols in official communication between ministries and agencies. Some of the target respondents especially among development partners were sent the questionnaires quite late in the data collection period which could have contributed to non-response or late response from those organizations. It was not clear why NIS was unable to convince key international development partners such as JICA and FAO, which have been among the major partners of NIS for many years, to participate in the study. Also, despite repeated requests and reminders from the consultants, the Country Team was unable to identify the appropriate focal persons and their contact information in some target respondents like ADB, USAID, and EU which have supported the statistical system in the past or recent years. At some points during the data collection, the international consultant and PARIS21 raised the need for high level intervention to increase participation in the study. There was no information if such measure was adopted. • Unclear role of the SCC in the CRESS

26

At the onset of the exercise, the CRESS design called for a steering committee to provide strategic directions and guidance and to enjoin ministries and agencies to cooperate. The existing SCC was specifically identified in the concept note for CRESS Cambodia which was accepted by NIS to serve as steering committee. No information was provided as to whether the SCC did assume the role or at least discussed the CRESS.

• Difficulty in collecting data from appropriate source/s within a ministry or agency

- Lack of coordination and clarity on the responsibility centers for accomplishing the

questionnaire

It was assumed that the designated statistical units were to accomplish the questionnaire for the ministries and agencies and to coordinate with other departments within the ministry similarly undertaking statistical activities, if any, as well as with the finance department/unit that handles budget information. However, in many of the ministries and agencies covered by the study, there was still confusion or lack of clarity as to which unit should fill out the questionnaires, which unit should coordinate with other units, or which units other than the designated statistics unit should be requested to provide additional data. Even in the MOP where there are 4 other directorates, in addition to NIS, it was not clear which department should coordinate the response of the ministry. Hoping to facilitate the process, the country team agreed to request all the directorates, except finance, to accomplish and submit the questionnaires. Eventually though, only 2 directorates were able to return the questionnaire and with very little input. Furthermore, NIS technically did not submit the accomplished questionnaires as it was the national consultant that filled them out for NIS based on an existing budget monitoring document earlier shared by the Secretary of State for statistics and some estimation done in consultation with the rest of the country team. Another example where data coordination proved difficult is the Ministry of Economy and Finance, which has administrative supervision of the General Department of Customs and Excise, the institution that compiles the country’s international merchandise statistics. - Inaccessibility of budget data and information

In most ministries/agencies, the designated statistical unit does not have access to budget and human resources data or at least the details thereof. In many cases, the budget is said to be embedded in bigger or ‘mother’ departments. In the case of NIS, which is under the MOP, there was clearly a huge challenge in getting data as financial management and information are centralized in the finance department. NIS does not have autonomy in budgetary and financial management. Higher level intervention could have facilitated the provision of budget data and information or the compilation of data from other sources such as the functional units or sectoral departments of NIS.

- Difficulty by government respondents in understanding the questionnaires and providing the

right data

27

Many of the respondents seemed to find the questionnaires difficult to understand and fill up even with the orientation, guidance, and instructions provided and the Khmer version made available and despite the explanation by the country team during follow up communication and personal visits. More specifically, there was difficulty identifying which staff positions or activities could be considered as statistical and the corresponding budget or expenditures. Some of the specific factors observed were:

There is no line item for statistics in the budget of many ministries and agencies, including

MOP. It is perhaps the same reason why NIS was unable to provide details of its budget – there is no separate line item for NIS in the MOP budget.

Statistical functions and activities of most ministries and agencies are embedded in bigger

departments/units, mostly planning, research, monitoring and evaluation and/or are undertaken only as part of non-statistician personnel’s duties. Only a handful of ministries and agencies have dedicated or separate statistical departments or units while some have only recently established their statistical units or the staff assigned to accomplish the questionnaires were new and did not have the relevant information.

It was not fully understood that statisticians include ICT personnel handling databases,

personnel in provincial or local offices, and analysts and researchers collecting and analyzing data for internal policy and decision making purposes of the ministry and agency.

It was difficult for ministries and agencies to identify which administrative systems or

research initiatives could be considered as statistical. It was difficult to identify the statistical component or activities and extract the

corresponding budget in projects supported by international development partners. Many respondents found it difficult to classify their statistical projects using the UNECE CSA.

The Country Team accepted the suggestion to make the questions on statistical domains optional for the respondents and agreed that the international consultant did the post-classification of activities. As it turned out, the post-classification proved challenging as well because many of the statistics projects address multi-subject areas hence the overlaps.

There was perception that most respondents had little knowledge about estimation

techniques and how these can help measure certain expenditure for statistics. Respondents were given more focused orientation on the different scenarios of statistics operations in ministries and agencies and the options by which expenditure or investment for statistics could be estimated. There was no feedback nor any indication that such orientation facilitated the accomplishment of the questionnaires.

One agency could only provide information on budget for personnel due to fiscal and

administrative autonomy and confidentiality issues.

• Inconsistencies of information across sources

28

There are significant discrepancies and inconsistencies between data on budget and expenditure for designated statistics published in the Annual Report of the Cambodia Statistical System and the responses of the relevant ministries and agencies to the CRESS, including NIS. Most of the representatives of the ministries and agencies that were shown the data were either not aware or not sure how the figures reflected in the said annual reports were arrived at. A quick assessment of the annual report data pointed to obvious quality issues. There is huge discrepancy between financing data reported by government and international development partners and among development partners. There are also observed differences between the responses of development partners in the CRESS study and PRESS reports. Responses from development partners indicate that the MOP and MOH were among those that received the highest levels of financial assistance for statistics-related projects. The assistance received by MOP was almost at par with what NIS received during the 2013-2016 period. In the case of MOH, there were many statistics projects reported by development partners that were not cited in the accomplished questionnaires. There is evidence that some statistics projects were implemented under a co-financing arrangement between ministries, between development partners, and between ministries and development partners. Responses though show inconsistencies. There was little option for validation because of, among other reasons, late submission of questionnaires and non-response from other involved development partners. Among the possible reasons for the inconsistencies are that: - The staff that accomplished the questionnaire was not aware of the statistics projects and

financing assistance. - The staff that accomplished the questionnaire did not have access to information on statistics

projects and financing assistance. - The staff that accomplished the questionnaire was new in the ministry/agency or unit and did

not have or had little knowledge about the statistics projects and budget. - The staff that accomplished the questionnaire was unable to coordinate with or collect the data

from the unit/s or person/s in charge of the statistics projects and financing assistance. - There was no coordination between the designated statistics unit and the finance

department/unit. - There was no coordination between the recipient ministries and agencies and the development

partners that provided the financing assistance. - There was no coordination between the country and global/main offices of development

partners. - There was weak coordination between development partners on co-financing initiatives.

One development partner that did not submit accomplished questionnaires lamented the poor documentation of financial assistance to government and transfer of records and knowledge between outgoing and incoming staff.

• Difficulty in analyzing financing of statistics in support of monitoring the designated official

statistics, and the NSDP and SDG indicators

29

A resource gap analysis would have been a significant output of the CRESS. Such analysis can show which of the designated official statistics, or NSDP and SDG indicators had been receiving financial support for development or improvement and which indicators were not being addressed or will be supported in the future. Limitations of the data however did not allow for the mapping of the designated official statistics, and NSDP and SDG indicators with past and future statistical initiatives of both the government and development partners.

Special note on the timing of the study There were several official and public holidays during the timetable of the project which made it difficult for respondents to accomplish the questionnaire and for the country team to make follow ups.

5. Lessons There have been many lessons learned during the exercise that will need to be revisited to guide follow up or future initiatives. • More interventions were needed to prepare the country team and respondents in accomplishing

the questionnaire. The launching forum is deemed insufficient as the participants may not necessarily be the same staff that will accomplish the questionnaires. A series of small group workshops according to clusters of ministries and agencies might help significantly in better appreciation of specific concepts and principles such as statistical functions and activities, statisticians or statistical personnel, statistical support, basic accounting concepts including expense categories, the budget law, etc. as well as learning about data management, including estimation techniques. The workshops could focus on real situations in the ministries and agencies to facilitate immediate solutions to issues and more guided accomplishment of the questionnaires. More importantly, such inputs could go a long way in building capacities to sustain the CRESS in the future. A separate coordination meeting among development partners could have contributed to higher response rate and more harmonized responses especially on statistics projects implemented on a co- financing scheme.

• The timetable for the CRESS should have taken into consideration the prevailing work practices (and

ethics) in the country as well as the technical and absorption capacity of the lead agency to manage the process. The numerous official and public holidays were hard to ignore as a factor in the timely collection of information not only in terms of time lost but also in how the delays affected momentum towards completion of the exercise. Because of delays in the data collection, the all so important data validation process had been limited. Careful planning is important but commitment and giving real priority to implementing the plan are more critical. A more collectively stable, well-trained, and committed country team could have minimized the challenges and ensured timely completion of outputs.

30

• Long-time partnership is no assurance of cooperation. The non-response of major development partners of NIS like JICA and FAO was unexpected, peculiar, and interesting at the same time. The participation of WHO and Australia/DFAT which provided information more than what had been reported in the PRESS and what had not been prominently mentioned in past documents was a welcome development.

• The CRESS could also capture support to the country that is part of global, regional, or multi-country

initiatives such as programs and projects implemented under the ASEAN Community Statistical System, training programs provided by UNSIAP and IMF, and regional economic and technical assistance of ADB, among others. Ways for the Country Team to communicate with these organizations should be explored.

• More effort to identify respondents beyond the ministries and agencies with designated statistics

could help determine support to statistics by the private sector, academic and research community, and non-government organizations.

6. Recommendations The CRESS Cambodia initiative has highlighted many of the usual key advocacy areas in statistics such as increasing appreciation for statistics, importance of intra- and inter-institutional coordination, clear delineation of roles, and strong leadership, and the need for continuous capacity building. The following recommendations are limited to those measures that can be considered or explored towards improving the CRESS as a tool to measure and monitor support for statistics towards improved statistical budgeting and financing in the country. A more rational, well-coordinated, and results-based approach to financing statistics will ensure that priority indicators and statistics are produced to support better analysis and comprehensive monitoring of the NSDP and SDGs, including progress in the continued development of the national statistical system. • NIS as the coordinator of the national statistical system may have to push for a separate budget

that can be managed independently from other directorates in the MOP. Further, since MOP covers a wide range of development concerns, including SDGs and poverty monitoring, that may require statistical studies and tools development, NIS may have to recommend ways for better record keeping and more coordinated planning, monitoring and reporting of statistical activities and projects within the MOP. These measures will facilitate better expenditure planning, program prioritization, resource mobilization, and monitoring of statistical initiatives.

• NIS may wish to consider learning more about the current government policy on program

budgeting and exploring avenues to integrate expenditure planning and monitoring for statistics. NIS may need high level intervention and support for closer collaboration with the Ministry of Economy and Finance on a concrete policy for statistics budgeting and monitoring of government allocation for statistics. One concrete goal is ensuring a separate line item for statistics in the budget of the ministries and agencies.

• The Cambodia statistical system may wish to consider CRESS as one of the mechanisms to help

monitor statistical development in the country. This can be done together with the preparation

31

of a wider monitoring and evaluation (M&E) plan for statistics that should include the monitoring of the NSDS, NSDP, and SDG indicators.

• Undertaking the CRESS requires adequate and appropriate technical capacity as well as high level support and intervention to ensure good coverage and participation of the stakeholders. The CRESS therefore needs to be included in strategic and technical discussions in the SCC and reported at the SAC or the Council of Ministers to gain more traction in terms of the policy solutions needed to address issues. NIS as coordinator of the national statistical system needs to take clear ownership and stronger leadership in pushing the envelope at the highest relevant political decision centers. It is strongly recommended that the CRESS be integrated into the functions or outputs of the NIS, together with the NSDS, with designated staff and resources to ensure continuity and further improvement.

• To promote increased participation among development partners, the CRESS should be included in regular or periodic donor coordination mechanisms including that between the NIS and development partners.

• For improved coverage and more complete and consistent financing data and information, - ministries and agencies need to strengthen the role of the designated statistical unit in

coordinating with other departments that carry out statistical activities or operations and to ensure better access to data and information on statistical development initiatives, including cooperation/collaborative projects with other ministries and agencies and development partners; NIS in particular may need to propose a mechanism whereby all statistics-related initiatives of the other directorates of MOP indicate clearly the role and contribution of NIS and the financing attributed to NIS.

- development partners need to improve documentation, ensure better transfer and sharing of information resources, and strengthen coordination with their regional and global offices, and with other development partners regarding support for statistics in the country;

- stronger coordination between ministries and agencies and their development partners in

managing financing data and information is essential; and - NIS, as coordinator of the national statistical system, should explore means to establishing

coordination with regional and global institutions such as UN Statistics Division, UNESCAP, UNSIAP, PARIS21 Secretariat, ADB, IMF, and ASEAN Secretariat in identifying and monitoring of financing support for statistics for Cambodia.

• NIS needs to take the lead in the further study and improvement of the CRESS framework and

methodology if regular monitoring of investment in statistics is considered a real priority. Initiatives to improve the CRESS should address the following areas:

32

- Coverage. The CRESS should: (a) increase response rate among and improve completeness of data from ministries and agencies with designated official statistics and development partners; and (b) cover expenditure for statistics of other ministries and agencies with no designated official statistics such as the MOP, National Training Board, National Employment Agency, local government units, academic and research institutions, non-government organizations, and the private sector, and regional/global development organizations. More details should be explored to allow for monitoring of the designated official statistics, NSDS, NSDP, and SDGs.

- Classification. NIS as the leading national statistics agency should improve understanding of the UNECE Classification of Statistical Activities and train statistical personnel in the ministries and agencies to promote harmonization and facilitate analysis of financing data by domain.

- Knowledge of the budgeting process. Understanding of the budget and the processes involved can facilitate the collection and reporting of financing data. Training of and learning sessions and discussions between statistical managers and finance officers on budgeting can contribute to improved budget planning and monitoring of statistics support.

- Consistency with the PRESS. The methodology could include a mechanism for improving

consistency of data between CRESS and PRESS.

• NIS should disseminate the CRESS results and findings to all stakeholders and initiate discussions of specific issues with relevant stakeholders. NIS should continue statistics advocacy for increased appreciation of and reasonable investment in statistics through the SCC, SAC, the Council of Ministers, the National Assembly, and other relevant forums.

33

Appendix

Appendix A – List of tables http://bit.ly/2iITWPJ

Appendix B – List of figures http://bit.ly/2iITWPJ

Appendix C – Results tables http://bit.ly/2AnhrEo

Appendix D - Respondents Information http://bit.ly/2BFFNrm

Appendix E – Acronyms http://bit.ly/2jbsKGx

Appendix F – List of designated official statistics http://bit.ly/2jbZjEb

Appendix G – References http://bit.ly/2A1dK8w

Annex Annex A – CRESS Cambodia Concept Note’

http://bit.ly/2Arldux (English) http://bit.ly/2zJK6Rb (Khmer)

Annex B – Launch programme, list of participants, and presentation materials http://bit.ly/2AEkRU1 (Programme) http://bit.ly/2BvQxI1 (Participants) http://bit.ly/2iJ4FtO (About the CRESS) http://bit.ly/2nt6tsu (Launch Objectives) http://bit.ly/2Annk4z (Methodology)

Annex C – Modified UNECE Classification of Statistical Activities http://bit.ly/2jbrMKp

Annex D – CRESS Cambodia Questionnaires http://bit.ly/2jIRkxN (English) http://bit.ly/2jGs2k9 (Khmer)

Annex E – CRESS Cambodia Database http://bit.ly/2BEVtev

Annex F – The CRESS Cambodia Country Team http://bit.ly/2zY7v5r

Annex G – Basic documents http://bit.ly/2iIcwHW (UNDP) http://bit.ly/2jbTtmt (SIDA) http://bit.ly/2jHfGYQ (UNICEF)